VMAF

Video Multi-Method Assessment Fusion, based on https://github.com/Netflix/vmaf.

Additionally, vsvmafxml can be used to store per-frame score from XML log as a frame property in clip.

Usage

vmaf.VMAF(vnode reference, vnode distorted, string log_path[, int log_format=0, int[] model=None, int[] feature=None])

• reference, distorted: Clips to compute VMAF score. Only YUV format with integer sample type of 8, 10, 12 and 16 bit depth and chroma subsampling of 420/422/444 is supported.

• log_path: Path to the log file.

• log_format: Format of the log file.

  • 0 = XML
  • 1 = JSON
  • 2 = CSV
  • 3 = subtitle

• model: Model to use. Refer to this, this and this page for more details.

  • 0 = vmaf_v0.6.1 (default mode)
  • 1 = vmaf_v0.6.1neg (NEG mode)
  • 2 = vmaf_b_v0.6.3 (Confidence Interval)
  • 3 = vmaf_4k_v0.6.1

• feature: Additional metrics to compute.

  • 0 = PSNR
  • 1 = PSNR-HVS
  • 2 = SSIM
  • 3 = MS-SSIM
  • 4 = CIEDE2000

---

vmaf.CAMBI(vnode clip, string log_path[, int log_format=0, int window_size=None, float topk=None, float tvi_threshold=None, int max_log_contrast=None, int enc_width=None, int enc_height=None])

CAMBI (Contrast Aware Multiscale Banding Index) is Netflix's detector for banding (aka contouring) artifacts. For an introduction to CAMBI, please refer to the tech blog.

The CAMBI score starts at 0, meaning no banding is detected. A higher CAMBI score means more visible banding artifacts are identified. The maximum CAMBI observed in a sequence is 24 (unwatchable). As a rule of thumb, a CAMBI score around 5 is where banding starts to become slightly annoying (also note that banding is highly dependent on the viewing environment - the brigher the display, and the dimmer the ambient light, the more visible banding is).

• clip: Clip to compute CAMBI score. Only YUV format with integer sample type of 8, 10, 12 and 16 bit depth and chroma subsampling of 420/422/444 is supported.

• log_path: Path to the log file.

• log_format: Format of the log file.

  • 0 = XML
  • 1 = JSON
  • 2 = CSV
  • 3 = subtitle

• window_size: (min: 15, max: 127, default: 63): Window size to compute CAMBI. (default: 63 corresponds to ~1 degree at 4K resolution and 1.5H)

• topk: (min: 0.0001, max: 1.0, default: 0.6): Ratio of pixels for the spatial pooling computation.

• tvi_threshold: (min: 0.0001, max: 1.0, default: 0.019): Visibilty threshold for luminance ΔL < tvi_threshold*L_mean for BT.1886.

• max_log_contrast: (min: 0, max: 5, default: 2): Maximum contrast in log luma level (2^max_log_contrast) at 10-bits. Default 2 is equivalent to 4 luma levels at 10-bit and 1 luma level at 8-bit. The default is recommended for banding artifacts coming from video compression.

• enc_width, enc_height: Encoding/processing resolution to compute the banding score, useful in cases where scaling was applied to the input prior to the computation of metrics.

---

vmaf.Metric(vnode reference, vnode distorted, int[] feature)

Compute the metrics and store the scores as frame properties.

• reference, distorted: Clips to compute score. Only YUV format with integer sample type of 8, 10, 12 and 16 bit depth and chroma subsampling of 420/422/444 is supported.

• feature: Metrics to compute.

  • 0 = PSNR
  • 1 = PSNR-HVS
  • 2 = SSIM
  • 3 = MS-SSIM
  • 4 = CIEDE2000

Compilation

Requires libvmaf.

meson build
ninja -C build
ninja -C build install